> 注意：此文件为机器翻译版本。欢迎对翻译进行改进！
English | 简体中文 | 日本語 | Português (Brasil) | 한국어

Comet Opik logo
Opik

开源 AI 可观测性、评估与优化平台

Opik 帮助您构建、测试并优化生成式 AI 应用，使其从原型到生产环境运行得更好。从 RAG 聊天机器人到代码助手再到复杂的智能体系统，Opik 提供全面的跟踪、评估，以及自动化的提示与工具优化，消除 AI 开发中的猜测。

[![Python SDK](https://img.shields.io/pypi/v/opik)](https://pypi.org/project/opik/) [![License](https://img.shields.io/github/license/comet-ml/opik)](https://github.com/comet-ml/opik/blob/main/LICENSE) [![Build](https://github.com/comet-ml/opik/actions/workflows/build_apps.yml/badge.svg)](https://github.com/comet-ml/opik/actions/workflows/build_apps.yml) [![Bounties](https://img.shields.io/endpoint?url=https%3A%2F%2Falgora.io%2Fapi%2Fshields%2Fcomet-ml%2Fbounties%3Fstatus%3Dopen)](https://algora.io/comet-ml/bounties?status=open)

官网 • Slack 社区 • Twitter • 更新日志 • 文档

🚀 什么是 Opik？ • 🛠️ Opik 服务端安装 • 💻 Opik 客户端 SDK • 📝 日志跟踪与集成
🧑‍⚖️ 作为裁判的 LLM • 🔍 评估您的应用 • ⭐ 在 GitHub 上给我们加星 • 🤝 贡献指南

[![Opik platform screenshot (thumbnail)](readme-thumbnail-new.png)](https://www.comet.com/signup?from=llm&utm_source=opik&utm_medium=github&utm_content=readme_banner&utm_campaign=opik) ## 🚀 什么是 Opik？ Opik（由 [Comet](https://www.comet.com?from=llm&utm_source=opik&utm_medium=github&utm_content=what_is_opik_link&utm_campaign=opik) 开发）是一款开源平台，旨在简化整个 LLM 应用生命周期。它让开发者能够评估、测试、监控和优化模型及智能体系统。主要功能包括： - **全面可观测性**：深度跟踪 LLM 调用、对话日志及智能体活动。 - **高级评估**：强大的提示评估、LLM-as-a-judge 及实验管理。 - **生产就绪**：可扩展的监控仪表板和在线评估规则。 - **Opik Agent Optimizer**：用于提升提示和智能体的专用 SDK 与优化器。 - **Opik Guardrails**：帮助您实施安全且负责任的 AI 实践。
主要功能包括： - **开发与跟踪:** - 在开发和生产环境中跟踪所有 LLM 调用和详细跟踪信息 ([快速开始](https://www.comet.com/docs/opik/quickstart/?from=llm&utm_source=opik&utm_medium=github&utm_content=quickstart_link&utm_campaign=opik)) - 丰富的第三方集成：原生支持 Google ADK、Autogen、Flowise AI 等主流框架 ([集成列表](https://www.comet.com/docs/opik/integrations/overview/?from=llm&utm_source=opik&utm_medium=github&utm_content=integrations_link&utm_campaign=opik)) - 通过 [Python SDK](https://www.comet.com/docs/opik/tracing/annotate_traces/#annotating-traces-and-spans-using-the-sdk?from=llm&utm_source=opik&utm_medium=github&utm_content=sdk_link&utm_campaign=opik) 或 [UI](https://www.comet.com/docs/opik/tracing/annotate_traces/#annotating-traces-through-the-ui?from=llm&utm_source=opik&utm_medium=github&utm_content=ui_link&utm_campaign=opik) 为跟踪和跨度添加反馈分数注释 - 在 [Prompt Playground](https://www.comet.com/docs/opik/prompt_engineering/playground) 中试验提示和模型 - **评估与测试**: - 使用 [Datasets](https://www.comet.com/docs/opik/evaluation/manage_datasets/?from=llm&utm_source=opik&utm_medium=github&utm_content=datasets_link&utm_campaign=opik) 和 [Experiments](https://www.comet.com/docs/opik/evaluation/evaluate_your_llm/?from=llm&utm_source=opik&utm_medium=github&utm_content=eval_link&utm_campaign=opik) 自动化 LLM 应用评估 - 利用 LLM-as-a-judge 指标进行复杂任务评估，如 [幻觉检测](https://www.comet.com/docs/opik/evaluation/metrics/hallucination/?from=llm&utm_source=opik&utm_medium=github&utm_content=hallucination_link&utm_campaign=opik)、[内容审核](https://www.comet.com/docs/opik/evaluation/metrics/moderation/?from=llm&utm_source=opik&utm_medium=github&utm_content=moderation_link&utm_campaign=opik) 和 RAG 评估（[回答相关性](https://www.comet.com/docs/opik/evaluation/metrics/answer_relevance/?from=llm&utm_source=opik&utm_medium=github&utm_content=alex_link&utm_campaign=opik)、[上下文精确度](https://www.comet.com/docs/opik/evaluation/metrics/context_precision/?from=llm&utm_source=opik&utm_medium=github&utm_content=context_link&utm_campaign=opik)） - 使用 [PyTest 集成](https://www.comet.com/docs/opik/testing/pytest_integration/?from=llm&utm_source=opik&utm_medium=github&utm_content=pytest_link&utm_campaign=opik) 将评估纳入 CI/CD 流水线 - **生产监控与优化**: - 高吞吐量生产跟踪：支持每日 4,000 万+ 跟踪记录 - 在 [Opik 仪表板](https://www.comet.com/docs/opik/production/production_monitoring/?from=llm&utm_source=opik&utm_medium=github&utm_content=dashboard_link&utm_campaign=opik) 中监控反馈分数、跟踪计数和令牌使用量 - 使用 [在线评估规则](https://www.comet.com/docs/opik/production/rules/?from=llm&utm_source=opik&utm_medium=github&utm_content=dashboard_link&utm_campaign=opik) 和 LLM-as-a-Judge 指标检测生产问题 - 利用 **Opik Agent Optimizer** 和 **Opik Guardrails** 持续改进和保护您的 LLM 应用 > [!TIP] > 如果您需要 Opik 当前尚不支持的功能，请提交新的 [功能请求](https://github.com/comet-ml/opik/issues/new/choose) 🚀
## 🛠️ Opik 服务端安装 几分钟内即可运行 Opik 服务端，选择最适合您的方案： ### 方案 1：Comet.com 云（最简易 & 推荐） 无需维护，立即体验 Opik。适合快速启动和无忧维护。 👉 [创建免费 Comet 帐号](https://www.comet.com/signup?from=llm&utm_source=opik&utm_medium=github&utm_content=install_create_link&utm_campaign=opik) ### 方案 2：自托管（完全掌控） 在您自己的环境中部署 Opik，本地开发可选 Docker Compose，大规模生产推荐 Kubernetes & Helm。 #### Docker Compose（本地开发 & 测试） 最简方式启动本地 Opik 实例，使用全新 `.opik.sh` 安装脚本： On Linux or Mac Environment: ```bash # 克隆 Opik 仓库 git clone https://github.com/comet-ml/opik.git # 进入仓库目录 cd opik # 启动 Opik 平台 ./opik.sh ``` On Windows Environment: ```powershell # 克隆 Opik 仓库 git clone https://github.com/comet-ml/opik.git # 进入仓库目录 cd opik # 启动 Opik 平台 powershell -ExecutionPolicy ByPass -c ".\\opik.ps1" ``` **开发服务配置文件** Opik 安装脚本现在支持针对不同开发场景的服务配置文件： ```bash # 完整 Opik 套件（默认行为） ./opik.sh # 仅基础设施服务（数据库、缓存等） ./opik.sh --infra # 基础设施 + 后端服务 ./opik.sh --backend # 在任何配置文件中启用守护栏 ./opik.sh --guardrails # 完整 Opik 套件 + 守护栏 ./opik.sh --backend --guardrails # 基础设施 + 后端 + 守护栏 ``` 使用 `--help` 或 `--info` 查看更多选项。Dockerfile 已确保容器以非 root 用户运行以增强安全性。启动成功后，打开浏览器访问 [localhost:5173](http://localhost:5173)。详情请见 [本地部署指南](https://www.comet.com/docs/opik/self-host/local_deployment?from=llm&utm_source=opik&utm_medium=github&utm_content=self_host_link&utm_campaign=opik)。 #### Kubernetes & Helm（大规模生产） 适用于生产或大规模自托管场景，通过 Helm Chart 在 Kubernetes 集群中安装 Opik： [![Kubernetes](https://img.shields.io/badge/Kubernetes-%23326ce5.svg?&logo=kubernetes&logoColor=white)](https://www.comet.com/docs/opik/self-host/kubernetes/#kubernetes-installation?from=llm&utm_source=opik&utm_medium=github&utm_content=kubernetes_link&utm_campaign=opik) > [!IMPORTANT] > **版本 1.7.0 变更**：请查看 [更新日志](https://github.com/comet-ml/opik/blob/main/CHANGELOG.md) 以了解重要更新和破坏性变更。 ## 💻 Opik 客户端 SDK Opik 提供一系列客户端库和 REST API 与 Opik 服务端交互，包含 Python、TypeScript 和 Ruby（通过 OpenTelemetry）SDK，方便集成到各类工作流中。详细 API 与 SDK 参考见 [客户端参考文档](https://www.comet.com/docs/opik/reference/overview?from=llm&utm_source=opik&utm_medium=github&utm_content=reference_link&utm_campaign=opik)。 ### Python SDK 快速开始 安装包： ```bash # 使用 pip 安装 pip install opik # 或使用 uv 安装 uv pip install opik ``` 运行 `opik configure`，并按提示输入 Opik 服务端地址（自托管）或 API key 与 workspace（Comet.com）： ```bash opik configure ``` > [!TIP] > 您也可以在代码中调用 `opik.configure(use_local=True)` 来配置本地自托管，或直接在代码中提供 API key 和 workspace。更多配置选项请参阅 [Python SDK 文档](https://www.comet.com/docs/opik/python-sdk-reference/?from=llm&utm_source=opik&utm_medium=github&utm_content=python_sdk_docs_link&utm_campaign=opik)。 现在您可以使用 [Python SDK](https://www.comet.com/docs/opik/python-sdk-reference/?from=llm&utm_source=opik&utm_medium=github&utm_content=sdk_link2&utm_campaign=opik) 记录跟踪！ ### 📝 日志跟踪与集成 最简单的跟踪方式是使用直接集成，Opik 支持多种框架，包括 Google ADK、Autogen、AG2 和 Flowise AI 等： | 集成 | 描述 | 文档 | | ------------------------- | ----------------------------------------------- | --------------------------------------------------------------------------------------------------------------------------------------------------------------------- | | **ADK** | 记录 Google Agent Development Kit (ADK) 的跟踪 | [文档](https://www.comet.com/docs/opik/integrations/adk?utm_source=opik&utm_medium=github&utm_content=google_adk_link&utm_campaign=opik) | | **AG2** | 记录 AG2 LLM 调用的跟踪 | [文档](https://www.comet.com/docs/opik/integrations/ag2?utm_source=opik&utm_medium=github&utm_content=ag2_link&utm_campaign=opik) | | **aisuite** | 记录 aisuite LLM 调用的跟踪 | [文档](https://www.comet.com/docs/opik/integrations/aisuite?utm_source=opik&utm_medium=github&utm_content=aisuite_link&utm_campaign=opik) | | **Agno** | 记录 Agno 智能体编排框架调用的跟踪 | [文档](https://www.comet.com/docs/opik/integrations/agno?utm_source=opik&utm_medium=github&utm_content=agno_link&utm_campaign=opik) | | **Anthropic** | 记录 Anthropic LLM 调用的跟踪 | [文档](https://www.comet.com/docs/opik/integrations/anthropic?utm_source=opik&utm_medium=github&utm_content=anthropic_link&utm_campaign=opik) | | **Autogen** | 记录 Autogen 智能体工作流的跟踪 | [文档](https://www.comet.com/docs/opik/integrations/autogen?utm_source=opik&utm_medium=github&utm_content=autogen_link&utm_campaign=opik) | | **Bedrock** | 记录 Amazon Bedrock LLM 调用的跟踪 | [文档](https://www.comet.com/docs/opik/integrations/bedrock?utm_source=opik&utm_medium=github&utm_content=bedrock_link&utm_campaign=opik) | | **BeeAI (Python)** | 记录 BeeAI Python 智能体框架调用的跟踪 | [文档](https://www.comet.com/docs/opik/integrations/beeai?utm_source=opik&utm_medium=github&utm_content=beeai_link&utm_campaign=opik) | | **BeeAI (TypeScript)** | 记录 BeeAI TypeScript 智能体框架调用的跟踪 | [文档](https://www.comet.com/docs/opik/integrations/beeai-typescript?utm_source=opik&utm_medium=github&utm_content=beeai_typescript_link&utm_campaign=opik) | | **BytePlus** | 记录 BytePlus LLM 调用的跟踪 | [文档](https://www.comet.com/docs/opik/integrations/byteplus?utm_source=opik&utm_medium=github&utm_content=byteplus_link&utm_campaign=opik) | | **CrewAI** | 记录 CrewAI 调用的跟踪 | [文档](https://www.comet.com/docs/opik/integrations/crewai?utm_source=opik&utm_medium=github&utm_content=crewai_link&utm_campaign=opik) | | **Cloudflare Workers AI** | 记录 Cloudflare Workers AI 调用的跟踪 | [文档](https://www.comet.com/docs/opik/integrations/cloudflare-workers-ai?utm_source=opik&utm_medium=github&utm_content=cloudflare_workers_ai_link&utm_campaign=opik) | | **Cohere** | 记录 Cohere LLM 调用的跟踪 | [文档](https://www.comet.com/docs/opik/integrations/cohere?utm_source=opik&utm_medium=github&utm_content=cohere_link&utm_campaign=opik) | | **Cursor** | 记录 Cursor 对话的跟踪 | [文档](https://www.comet.com/docs/opik/integrations/cursor?utm_source=opik&utm_medium=github&utm_content=cursor_link&utm_campaign=opik) | | **DeepSeek** | 记录 DeepSeek LLM 调用的跟踪 | [文档](https://www.comet.com/docs/opik/integrations/deepseek?utm_source=opik&utm_medium=github&utm_content=deepseek_link&utm_campaign=opik) | | **Dify** | 记录 Dify 智能体运行的跟踪 | [文档](https://www.comet.com/docs/opik/integrations/dify?utm_source=opik&utm_medium=github&utm_content=dify_link&utm_campaign=opik) | | **DSPy** | 记录 DSPy 运行的跟踪 | [文档](https://www.comet.com/docs/opik/integrations/dspy?utm_source=opik&utm_medium=github&utm_content=dspy_link&utm_campaign=opik) | | **Flowise AI** | 记录 Flowise AI 可视化 LLM 应用的跟踪 | [文档](https://www.comet.com/docs/opik/integrations/flowise?utm_source=opik&utm_medium=github&utm_content=flowise_link&utm_campaign=opik) | | **Fireworks AI** | 记录 Fireworks AI LLM 调用的跟踪 | [文档](https://www.comet.com/docs/opik/integrations/fireworks-ai?utm_source=opik&utm_medium=github&utm_content=fireworks_ai_link&utm_campaign=opik) | | **Gemini (Python)** | 记录 Google Gemini LLM 调用的跟踪 | [文档](https://www.comet.com/docs/opik/integrations/gemini?utm_source=opik&utm_medium=github&utm_content=gemini_link&utm_campaign=opik) | | **Gemini (TypeScript)** | 记录 Google Gemini TypeScript SDK 调用的跟踪 | [文档](https://www.comet.com/docs/opik/integrations/gemini-typescript?utm_source=opik&utm_medium=github&utm_content=gemini_typescript_link&utm_campaign=opik) | | **Groq** | 记录 Groq LLM 调用的跟踪 | [文档](https://www.comet.com/docs/opik/integrations/groq?utm_source=opik&utm_medium=github&utm_content=groq_link&utm_campaign=opik) | | **Guardrails** | 记录 Guardrails AI 验证的跟踪 | [文档](https://www.comet.com/docs/opik/integrations/guardrails-ai?utm_source=opik&utm_medium=github&utm_content=guardrails_link&utm_campaign=opik) | | **Haystack** | 记录 Haystack 调用的跟踪 | [文档](https://www.comet.com/docs/opik/integrations/haystack?utm_source=opik&utm_medium=github&utm_content=haystack_link&utm_campaign=opik) | | **Harbor** | 记录 Harbor 基准评估试验的跟踪 | [文档](https://www.comet.com/docs/opik/integrations/harbor?utm_source=opik&utm_medium=github&utm_content=harbor_link&utm_campaign=opik) | | **Instructor** | 记录 Instructor LLM 调用的跟踪 | [文档](https://www.comet.com/docs/opik/integrations/instructor?utm_source=opik&utm_medium=github&utm_content=instructor_link&utm_campaign=opik) | | **LangChain (Python)** | 记录 LangChain LLM 调用的跟踪 | [文档](https://www.comet.com/docs/opik/integrations/langchain?utm_source=opik&utm_medium=github&utm_content=langchain_link&utm_campaign=opik) | | **LangChain (JS/TS)** | 记录 LangChain JavaScript/TypeScript 调用的跟踪 | [文档](https://www.comet.com/docs/opik/integrations/langchainjs?utm_source=opik&utm_medium=github&utm_content=langchainjs_link&utm_campaign=opik) | | **LangGraph** | 记录 LangGraph 执行的跟踪 | [文档](https://www.comet.com/docs/opik/integrations/langgraph?utm_source=opik&utm_medium=github&utm_content=langgraph_link&utm_campaign=opik) | | **Langflow** | 记录 Langflow 可视化 AI 应用的跟踪 | [文档](https://www.comet.com/docs/opik/integrations/langflow?utm_source=opik&utm_medium=github&utm_content=langflow_link&utm_campaign=opik) | | **LiteLLM** | 记录 LiteLLM 模型调用的跟踪 | [文档](https://www.comet.com/docs/opik/integrations/litellm?utm_source=opik&utm_medium=github&utm_content=litellm_link&utm_campaign=opik) | | **LiveKit Agents** | 记录 LiveKit Agents AI 智能体框架调用的跟踪 | [文档](https://www.comet.com/docs/opik/integrations/livekit?utm_source=opik&utm_medium=github&utm_content=livekit_link&utm_campaign=opik) | | **Mastra** | 记录 Mastra AI 工作流框架调用的跟踪 | [文档](https://www.comet.com/docs/opik/integrations/mastra?utm_source=opik&utm_medium=github&utm_content=mastra_link&utm_campaign=opik) | | **Microsoft Agent Framework (Python)** | 记录 Microsoft Agent Framework 调用的跟踪 | [文档](https://www.comet.com/docs/opik/integrations/microsoft-agent-framework?utm_source=opik&utm_medium=github&utm_content=agent_framework_link&utm_campaign=opik) | | **Microsoft Agent Framework (.NET)** | 记录 Microsoft Agent Framework .NET 调用的跟踪 | [文档](https://www.comet.com/docs/opik/integrations/microsoft-agent-framework-dotnet?utm_source=opik&utm_medium=github&utm_content=agent_framework_dotnet_link&utm_campaign=opik) | | **Mistral AI** | 记录 Mistral AI LLM 调用的跟踪 | [文档](https://www.comet.com/docs/opik/integrations/mistral?utm_source=opik&utm_medium=github&utm_content=mistral_link&utm_campaign=opik) | | **n8n** | 记录 n8n 工作流执行的跟踪 | [文档](https://www.comet.com/docs/opik/integrations/n8n?utm_source=opik&utm_medium=github&utm_content=n8n_link&utm_campaign=opik) | | **LlamaIndex** | 记录 LlamaIndex LLM 调用的跟踪 | [文档](https://www.comet.com/docs/opik/integrations/llama_index?utm_source=opik&utm_medium=github&utm_content=llama_index_link&utm_campaign=opik) | | **Ollama** | 记录 Ollama LLM 调用的跟踪 | [文档](https://www.comet.com/docs/opik/integrations/ollama?utm_source=opik&utm_medium=github&utm_content=ollama_link&utm_campaign=opik) | | **OpenAI (Python)** | 记录 OpenAI LLM 调用的跟踪 | [文档](https://www.comet.com/docs/opik/integrations/openai?utm_source=opik&utm_medium=github&utm_content=openai_link&utm_campaign=opik) | | **OpenAI (JS/TS)** | 记录 OpenAI JavaScript/TypeScript 调用的跟踪 | [文档](https://www.comet.com/docs/opik/integrations/openai-typescript?utm_source=opik&utm_medium=github&utm_content=openai_typescript_link&utm_campaign=opik) | | **OpenAI Agents** | 记录 OpenAI Agents SDK 调用的跟踪 | [文档](https://www.comet.com/docs/opik/integrations/openai_agents?utm_source=opik&utm_medium=github&utm_content=openai_agents_link&utm_campaign=opik) | | **Novita AI** | 记录 Novita AI LLM 调用的跟踪 | [文档](https://www.comet.com/docs/opik/integrations/novita-ai?utm_source=opik&utm_medium=github&utm_content=novita_ai_link&utm_campaign=opik) | | **OpenRouter** | 记录 OpenRouter LLM 调用的跟踪 | [文档](https://www.comet.com/docs/opik/integrations/openrouter?utm_source=opik&utm_medium=github&utm_content=openrouter_link&utm_campaign=opik) | | **OpenTelemetry** | 记录 OpenTelemetry 支持的调用跟踪 | [文档](https://www.comet.com/docs/opik/tracing/opentelemetry/overview?utm_source=opik&utm_medium=github&utm_content=opentelemetry_link&utm_campaign=opik) | | **OpenWebUI** | 记录 OpenWebUI 对话的跟踪 | [文档](https://www.comet.com/docs/opik/integrations/openwebui?utm_source=opik&utm_medium=github&utm_content=openwebui_link&utm_campaign=opik) | | **Pipecat** | 记录 Pipecat 实时语音智能体调用的跟踪 | [文档](https://www.comet.com/docs/opik/integrations/pipecat?utm_source=opik&utm_medium=github&utm_content=pipecat_link&utm_campaign=opik) | | **Predibase** | 记录 Predibase LLM 调用的跟踪 | [文档](https://www.comet.com/docs/opik/integrations/predibase?utm_source=opik&utm_medium=github&utm_content=predibase_link&utm_campaign=opik) | | **Pydantic AI** | 记录 PydanticAI 智能体调用的跟踪 | [文档](https://www.comet.com/docs/opik/integrations/pydantic-ai?utm_source=opik&utm_medium=github&utm_content=pydantic_ai_link&utm_campaign=opik) | | **Ragas** | 记录 Ragas 评估的跟踪 | [文档](https://www.comet.com/docs/opik/integrations/ragas?utm_source=opik&utm_medium=github&utm_content=ragas_link&utm_campaign=opik) | | **Smolagents** | 记录 Smolagents 智能体调用的跟踪 | [文档](https://www.comet.com/docs/opik/integrations/smolagents?utm_source=opik&utm_medium=github&utm_content=smolagents_link&utm_campaign=opik) | | **Semantic Kernel** | 记录 Microsoft Semantic Kernel 调用的跟踪 | [文档](https://www.comet.com/docs/opik/integrations/semantic-kernel?utm_source=opik&utm_medium=github&utm_content=semantic_kernel_link&utm_campaign=opik) | | **Spring AI** | 记录 Spring AI 框架调用的跟踪 | [文档](https://www.comet.com/docs/opik/integrations/spring-ai?utm_source=opik&utm_medium=github&utm_content=spring_ai_link&utm_campaign=opik) | | **Strands Agents** | 记录 Strands Agents 调用的跟踪 | [文档](https://www.comet.com/docs/opik/integrations/strands-agents?utm_source=opik&utm_medium=github&utm_content=strands_agents_link&utm_campaign=opik) | | **Together AI** | 记录 Together AI LLM 调用的跟踪 | [文档](https://www.comet.com/docs/opik/integrations/together-ai?utm_source=opik&utm_medium=github&utm_content=together_ai_link&utm_campaign=opik) | | **Vercel AI SDK** | 记录 Vercel AI SDK 调用的跟踪 | [文档](https://www.comet.com/docs/opik/integrations/vercel-ai-sdk?utm_source=opik&utm_medium=github&utm_content=vercel_ai_sdk_link&utm_campaign=opik) | | **VoltAgent** | 记录 VoltAgent 智能体框架调用的跟踪 | [文档](https://www.comet.com/docs/opik/integrations/voltagent?utm_source=opik&utm_medium=github&utm_content=voltagent_link&utm_campaign=opik) | | **watsonx** | 记录 IBM watsonx LLM 调用的跟踪 | [文档](https://www.comet.com/docs/opik/integrations/watsonx?utm_source=opik&utm_medium=github&utm_content=watsonx_link&utm_campaign=opik) | | **xAI Grok** | 记录 xAI Grok LLM 调用的跟踪 | [文档](https://www.comet.com/docs/opik/integrations/xai-grok?utm_source=opik&utm_medium=github&utm_content=xai_grok_link&utm_campaign=opik) | > [!TIP] > 如果您使用的框架不在上述列表中，请 [打开 Issue](https://github.com/comet-ml/opik/issues) 或提交 PR。 如果您未使用任何框架，也可以使用 `track` 装饰器记录跟踪（[详情](https://www.comet.com/docs/opik/tracing/log_traces/?from=llm&utm_source=opik&utm_medium=github&utm_content=traces_link&utm_campaign=opik)）： ```python import opik opik.configure(use_local=True) # 本地运行 @opik.track def my_llm_function(user_question: str) -> str: # 在此处编写您的 LLM 代码 return "你好" ``` > [!TIP] > `track` 装饰器可与任何集成结合使用，亦可用于跟踪嵌套函数调用。 ### 🧑‍⚖️ 作为裁判的 LLM Python Opik SDK 包含多种 LLM-as-a-judge 指标，可帮助您评估 LLM 应用。详情请参阅 [指标文档](https://www.comet.com/docs/opik/evaluation/metrics/overview/?from=llm&utm_source=opik&utm_medium=github&utm_content=metrics_2_link&utm_campaign=opik)。 使用示例： ```python from opik.evaluation.metrics import Hallucination metric = Hallucination() score = metric.score( input="法国的首都是哪里？", output="巴黎", context=["法国是欧洲的一个国家。"] ) print(score) ``` Opik 还提供多种预构建启发式指标，并支持创建自定义指标。更多信息请参阅同一 [指标文档](https://www.comet.com/docs/opik/evaluation/metrics/overview/?from=llm&utm_source=opik&utm_medium=github&utm_content=metrics_3_link&utm_campaign=opik)。 ### 🔍 评估您的应用 在开发过程中，可使用 [Datasets](https://www.comet.com/docs/opik/evaluation/manage_datasets/?from=llm&utm_source=opik&utm_medium=github&utm_content=datasets_2_link&utm_campaign=opik) 和 [Experiments](https://www.comet.com/docs/opik/evaluation/evaluate_your_llm/?from=llm&utm_source=opik&utm_medium=github&utm_content=experiments_link&utm_campaign=opik) 进行评估。Opik 仪表板提供增强的实验图表并改进大规模跟踪处理。您还可以使用 [PyTest 集成](https://www.comet.com/docs/opik/testing/pytest_integration/?from=llm&utm_source=opik&utm_medium=github&utm_content=pytest_2_link&utm_campaign=opik) 将评估纳入 CI/CD 流程。 ## ⭐ 在 GitHub 上给我们加星 如果您觉得 Opik 有用，请在 GitHub 上给我们加星！您的支持有助于我们壮大社区并持续改进产品。 [![Star History Chart](https://api.star-history.com/svg?repos=comet-ml/opik&type=Date)](https://github.com/comet-ml/opik) ## 🤝 贡献指南 贡献 Opik 的方式有很多： - 提交 [错误报告](https://github.com/comet-ml/opik/issues) 和 [功能请求](https://github.com/comet-ml/opik/issues) - 审阅文档并提交 [Pull Requests](https://github.com/comet-ml/opik/pulls) 改进文档 - 在演讲或文章中介绍 Opik 并[告诉我们](https://chat.comet.com) - 为热门 [功能请求](https://github.com/comet-ml/opik/issues?q=is%3Aissue+is%3Aopen+label%3A%22enhancement%22) 投票表示支持 更多详情请参阅 [CONTRIBUTING.md](CONTRIBUTING.md)。